We’ve been quietly working on something for a while now, and we’re finally ready to bring more people in.
Over the past few months, we’ve run a series of roundtables with community members, and we’ve been really grateful for everyone who came and shared their thoughts. That kind of engagement is exactly what makes building with this community so worthwhile, and a lot of what we heard is showing up in what we’ve been building. Now we’re taking the next step: opening up a private beta for a docs experience we’re genuinely excited about.
This is an early look: some things are still in progress, and not everything will be available right away. But the core of what we’ve been building is there, and we think you’re going to like where it’s headed.
When you join the waitlist, you’ll hear from our team with the next steps once we’re ready to bring you in.
Could you give a TLDR on what exactly Coda Docs might be, before I fill out the long form to join the wait list?
By the way some quick feedback: I’m using Coda less and less these days because I’m having trouble interacting with the API in Claude Code to read, adjust, propagate, etc. It can access tables and adjust them but it’s much harder to do anything with the pages content. So the tables are not really in a structured database, which makes code-based work harder and I’m trying to move towards a sql database format. (Probably not alone in this.)
So if Coda Docs helps with this, that would be music to my ears.
I agree with @Contact_Unjournal , can you please share some details? What are we signing up for? Your message almost reads like you meant to share some insights but then you hit Send too early.
@Contact_Unjournal , I doubt Coda will address your concerns. It’s not a real relational database in the way Supabase or Fibery would be. I don’t think Coda was intended to be a solid RDBMS.
btw - there’s mentions of the new Database in the forums if you go digging - but if you’re hoping to have a ‘Prisma Studio’ style live view of a production database inside Coda, I suspect your gonna be disappointed - i think the direction there taking is that Coda is the ‘article of truth’ - like they want all your data - and Coda is not unique in that approach re: AI - all agency AI like CoWork / OpenClaw style wants to be single source of truth.
How would we be able to query the super tables? Has an SQL type of querying been contemplated?
I’m concerned that the co-existence super tables and regular tables will cause unnecessary complexity and confusion. In the least, the user will have to decide - when do I use a regular table and when a super table. Would I be able to link a super table with a regular table? By having both supertables and regular tables, you enable the creation of two tiers of databases. I can see how this can be powerful but also extremely messy.
Following on point 2, wouldn’t it make more sense to have only supertables with entities as first class citizens callable across any doc in the workspace. Forget the regular tables. Supertables make them obsolete. For examples of how this is executed well, look at Fibery. Their bases and tables can be imported and referenced in any doc, dashboard, report and other tables all across the workspace.
I appreciate you sharing this before filling out the form. The feedback is genuinely useful.
Quick answer on Docs: it’s a new document experience we’re building inside Superhuman, designed to make collaborative docs feel faster and smarter. More details are coming as we get closer to launch, but the beta is a chance to get in early and help shape it.
On your API/Claude Code pain point: have you had a chance to try our MCP yet? It was built specifically to give AI agents like Claude Code a much cleaner way to interact with Coda content, including pages, not just tables. That page content access issue you’re describing is exactly what it’s meant to address, so it’s worth a try if you haven’t already.
That said, I want to be straight with you: If the deeper frustration is that Coda tables don’t behave like a relational SQL database, MCP improves the interaction layer but doesn’t change the underlying architecture. If SQL-style structure is the core need, I’d rather you know that going in than oversell what we’ve built.
Either way, your feedback is exactly the kind we want in the beta. Hope you’ll join us!
You raise a fair point, and I totally get it. The vagueness is intentional rather than an oversight. We’re in early beta specifically because the product is still being shaped by feedback, and I’d rather be upfront about that than publish a detailed spec that shifts on us two weeks from now.
Honestly, opinions like yours are exactly what we want in the room while we’re building this. More will become clear once you’re in.
And on the relational database point, you’re not wrong about what Coda is and isn’t. That kind of clear-eyed context is exactly what’s useful when someone is weighing whether to join.
Hi, I understand the need to keep this vague. On the other hand, to my imagination this reads more like you’re creating a new user experience, possibly changing the UI or basic functionality of docs / pages / canvases drastically, or maybe the organization of docs.
I know supertables have been discussed and I’m superexcited about it, but for me “a new document experience” doesn’t point to that.
I’m up for both changes but for the first I’m not ready to bring in the whole team and would rather sign up with my personal account. For the latter I would be interested to sign up the whole team.
@Niko_Nyman@Piet_Strydom — there will be some updates to the UI, more powerful AI features, etc… but it is not going to be a wholesale change that will greatly disrupt what you’ve already built. This is a “Yes, and…” kind of thing and not a “No, but…” The goal is to optimize, not disrupt. It would be great to get you both in so you can take a look for yourselves!
Can’t wait to get my hands on this exciting beta to see if Coda’s new workspace-wide DB experience is on par with Fibery’s boundless interconnectivity. So many super sweet things I’ve been waiting to build!
but for the first I’m not ready to bring in the whole team and would rather sign up with my personal account. For the latter I would be interested to sign up the whole team.
So many super sweet things I’ve been waiting to build! 